Star. Number of Latitude Observer.
Observations. deduced.
deg. min. sec.
alpha Lyrae 8 2 31 42.4 Capt. Sabine
alpha Lyrae 12 43.8 Ditto
alpha Pavonis 10 44.5 Ditto
alpha Lyrae 12 44.6 Ditto
alpha Cygni 12 42.1 Ditto
alpha Gruris 12 42.2 Ditto
Mean latitude deduced from 66 observations 2deg. 31min 43.3sec.
In comparing these results, although the French observations were more than twenty times as numerous as the English, yet the deviations of the individual sets from the mean are greater. One second and three-tenths is the greatest deviation from the mean of the Maranham observations; whilst the greatest deviation of those of Formentera, is two seconds and two-tenths. If this mode of comparison should be thought unfair, on account of the greater number of the sets in the French observations, let any six, in succession, of those sets be taken, and compared with the six English sets; and it will be found that in no one instance is the greatest deviation from the mean of the whole of the observations less than in those of Maranham. It must also be borne in mind, that by the latitude deduced by the mean of 1250 superior culminations of Polaris by the same observers, the latitude of Formentera was found to be 38deg. 39min 57.07sec., a result differing by 2.14sec. from the mean of the 1318 inferior culminations given above. [This difference cannot be accounted for by any difference in the tables of refraction, as neither the employment of those of Bradley, of Piazzi, of the French, of Groombridge, of Young, of Ivory, of Bessel, or of Carlini, would make a difference of two-tenths of a second.]
These facts alone ought to have awakened the attention of Captain Sabine, and of those who examined and officially pronounced on the merits of his observations; for, supposing the skill of the observers equal, it seems a necessary consequence that "the performance of the six-inch circle is" not merely "fully equal to that of circles of larger dimensions," but that it is decidedly SUPERIOR to one of sixteen inches in diameter.
This opinion did indeed gain ground for a time; but, fortunately for astronomy, long after these observations were made, published, and rewarded, Captain Kater, having borrowed the same instrument, discovered that the divisions of its level, which Captain Sabine had considered to be equal to one second each, were, in fact, more nearly equal to eleven seconds, each one being 10.9sec. This circumstance rendered necessary a recalculation of all the observations made with that instrument: a re-calculation which I am not aware Captain Sabine has ever thought it necessary to publish. [Above two hundred sets of observations with this instrument are given in the work alluded to. It can never be esteemed satisfactory merely to state the mean results of the corrections arising from this error: for the confidence to be attached to that mean will depend on the nature of the deviations from it.]
This is the more to be regretted, as it bears upon a point of considerable importance to navigation; and if it should have caused any alteration in his opinion as to the comparative merits of great and small instruments, it might have been expected from a gentleman, who was expressly directed by the Board of Longitude, to try the question with an instrument constructed for that especial purpose.
Finding that this has not been done by the person best qualified for the task, perhaps a few remarks from one who has no pretensions to familiarity with the instrument, may tend towards elucidating this interesting question.
The following table gives the latitudes as corrected for the error of level:
Station. Star Latitude Latitude Diffe-
by Capt. corrected for rence
Sabine error of level.
deg.min.sec. deg.min.sec. sec.
Sierra Leone Sirius 8 29 27.9 8 29 34.7 6.8
Ascension Alph.Centuri 7 55 46.7 7 55 40.1 6.6
Bahia Alph.Lyrae 12 59 19.4 12 59 21.4 2.0
Alph.Lyrae 21.2 58 49.8 31.4
Alph.Pavonis 22.4 59 5.1 17.3
Maranham Alph.Lyrae 2 31 42.4 2 31 22 20.4
Alph.Lyrae 43.8 31.8 12.0
Alph.Pavonis 44.5 44 .5
Alph.Lyrae 44.6 42.6 2.0
Alph.Cygni 42.1 39.2 2.9
Alph.Gruris 42.2 27.4 14.8
Trinidad Achernar 10 38 56.1 10 38 58.2 2.1
Alph.Gruris 52.2 50.8 1.4
Achernar 59.3 56.6 2.7
Jamaica Polaris 17 56 8.6 17 56 4.6 4.0
6.6 3.3 3.3
New York Sun 40 42 40.1 40 42 44.6 4.5
Polaris 48.9 38.2 10.7
Sun 41.4 47.2 5.8
Beta Urs.Min. 42.3 58.4 16.1
Hammerfest Sun 70 40 5.3 70 40 7.2 1.9
Spitzbergen Sun 79 49 56.1 79 49 58.6 2.5
Sun 55.9 44.8 11.1
Sun 58.6 52.7 5.9
Sun 59.3 51.6 7.7
Sun 55.8 51.6 4.2
Sun 50 1.5 57.0 4.5
Greenland Sun 74 32 19.9 74 32 32.4 12.4
Sun 17.9 18.7 0.8
Drontheim Sun 63 25 51.3 63 26 6.1 14.8
Alph.Urs.Min. 57.2 49.4 7.8
This presents a very different view of the latitudes as determined by the small repeating circle, from that in Captain Sabine's book; and confining ourselves still to Maranham, where the latitudes "WERE OBTAINED, WITH ESPECIAL REGARD TO EVERY CIRCUMSTANCE BY WHICH THEIR ACCURACY MIGHT BE AFFECTED," and where "A MORE THAN USUAL ATTENTION WAS BESTOWED," it appears, that if we take Captain Sabine's own test, namely, "the differences of the partial results from the mean at each station," the deviations become nearly ten times as large as they were before; a circumstance which might be expected to have some influence in the decision of the question.
There is, however, another light in which it is impossible to avoid looking at this singular oversight. The second column of the table of latitudes must now be considered the true one, as that which really resulted from the observations. Now, on examining the column of true latitudes, the differences between the different sets of observations is so considerable as naturally to excite some fear of latent error, more especially as nearly the greatest discordance arises from the same star, Alph.Lyrae, observed after an interval of only three days. It becomes interesting to every person engaged in making astronomical observations, to know what is the probability of his being exposed to an error so little to be guarded against, and so calculated to lull the suspicions of the unfortunate astronomer to whom it may happen.